Provide builtin impls of Fn traits for fn-pointers

[hban]

Meant to be, but isn't actually a fix for #2880.

Consider this snippet:

use std::marker::PhantomData;
use std::ops::Deref;

struct Lazy<T, F/* = fn() -> T*/>(F, PhantomData<T>);

impl<T, F> Lazy<T, F> {
    pub fn new(f: F) -> Lazy<T, F> {
        Lazy(f, PhantomData)
    }
}

impl<T, F: FnOnce() -> T> Deref for Lazy<T, F> {
    type Target = T;

    fn deref(&self) -> &T { todo!() }
}

fn test() {
    let lazy1: Lazy<u32, _> = Lazy::new(|| 0u32);
    let r1 = lazy1.to_string();

    fn make_u32_fn() -> u32 { todo!() }
    let make_u32_fn_ptr: fn() -> u32 = make_u32_fn;
    let lazy2: Lazy<u32, _> = Lazy::new(make_u32_fn_ptr);
    let r2 = lazy2.to_string();
}

• On current master:
  • When type default is commented-out, r1 is correctly inferred, r2 in {unknown}.
  • When type default is not commented-out, both r1 and r2 are {unknown}.
• With this PR:
  • When type default is commented-out, both r1 and r2 are correctly inferred.
  • When type default is not commented-out, both r1 and r2 are {unknown}.

Well, it's a improvement at least. I guess this thing with type defaults is a different problem.

I also tried add Fn impls for fn items, but wasn't successful. So this PR only adds those impls for fn pointers.

[flodiebold]

Thanks for your contribution! This is somewhat unfortunate though, both because there was another PR for the same thing just yesterday (#4483), and because the instructions in the bug were outdated -- Chalk is now providing built-in impls, so these should be added there as well: rust-lang/chalk#363.

I think at least the test would be good to keep though, even if it's just documenting how it's currently not working.

[hban]

Bad timing, indeed 😄

That "good to keep" test, you mean one added in this PR, or that snippet in my first comment?

[flodiebold]

Hm, both, actually!

[hban]

Done.

[flodiebold]

Thanks!

bors r+

[bors]

Build succeeded:

• Rust (macos-latest)
• Rust (ubuntu-latest)
• Rust (windows-latest)
• TypeScript